AUTHORS
Yuying Zhu, PhD, Brian John Ball, MD, MS2, Hyunjun Kang, PhD3, Melissa Valerio, PhD4, Seongsik Won5, Dinh Hoang6, Jessica Dang, PhD7, Amanda Blackmon, DO8, Zhuo Li, PhD9, Bin Zhang, PhD10, Jiange Zhang, PhD11, Guido Marcucci, MD12, Le Xuan Truong Nguyen, PhD13
AFFILIATIONS
1 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
2 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
3 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
4 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
5 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
6 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
7 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
8 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
9 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
10 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
11 Shanghai University of Traditional Chinese Medicine, Shanghai, China
12 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
13 Beckman Research Institute and City of Hope National Medical Center, Duarte, California, United States
ABSTRACT
Ferroptosis is an iron-dependent form of cell death driven by excessive lipid peroxidation and represents a promising strategy for eliminating therapy-resistant leukemic stem cells (LSCs) in acute myeloid leukemia (AML). However, LSCs evade ferroptosis through activation of nuclear factor erythroid 2-related factor 2 (NRF2)-dependent antioxidant programs. NRF2 regulates antioxidant and ferroptosis-protective genes, including SLC7A11, GPX4, and NQO1, promoting leukemia cell survival and therapeutic resistance. Gambogenic acid (GNA) was recently identified as a NRF2 inhibitor that promotes ubiquitin-mediated degradation of the NRF2/MafG complex and induces ferroptosis in NRF2-driven solid tumors. Whether GNA exploits NRF2-dependent ferroptosis to eliminate LSCs in AML remains unknown.
We show that GNA potently inhibited AML cell lines (IC₅₀,0.9–1.8μM) and primary CD34⁺CD38⁻ LSC-enriched AML blasts (IC₅₀,1.1–1.7μM), while sparing normal CD34⁺CD38⁻ hematopoietic stem/progenitor cells. At 1μM, GNA reduced colony formation of primary CD34⁺CD38⁻ AML cells by ~90% without affecting normal controls. RNA sequencing of GNA- and vehicle-treated MOLM-13 cells identified 380 differentially expressed genes. GSEA revealed activation of reactive oxygen species (ROS) signaling with suppression of oxidative phosphorylation (OXPHOS), suggesting disruption of redox homeostasis and mitochondrial metabolism. Consistently, GNA (0.5μM) suppressed fatty acid oxidation (FAO)/OXPHOS while increasing mitochondrial ROS and inducing ferroptosis, evidenced by increased lipid peroxidation (BODIPY and MDA) and decreased GPX4 activity in HL-60 cells and CD34⁺CD38⁻ AML blasts. Mechanistically, GNA promoted NRF2 ubiquitination and proteasomal degradation, reducing expression of the NRF2 downstream target involved in antioxidant defense (NQO1) and ferroptosis suppression (SLC7A11 and GPX4). NRF2 knockdown phenocopied GNA treatment, whereas NRF2 overexpression partially rescued the GNA-induced suppression of FAO/OXPHOS and induction of ferroptosis, establishing NRF2 as a critical mediator of GNA activity. In parallel, GNA reduced OPA1 expression, a key regulator of mitochondrial fusion, and decreased mitochondrial size (control, 1.0μm vs. GNA, 0.7μm), consistent with impaired mitochondrial fusion. These mitochondrial alterations were associated with suppression of FAO/OXPHOS and LSC elimination.
In vivo, GNA (6 mg/kg, IP, 2x/week for 3 weeks) significantly reduced leukemia burden in both MllPTD/WT/Flt3ITD/ITD transgenic and FLT3-WT PDX models. In MllPTD/WT/Flt3ITD/ITD mice, GNA reduced peripheral blood (PB) engraftment (6.9% vs. 21.5%, p<0.0001) and prolonged survival (median survival [MS],42 vs. 35 days, p <0.001). Comparable antileukemic activity was observed in FLT3-WT PDX mice, with reduced engraftment (18.8% vs. 35.5%, p<0.001) and prolonged survival (MS, 54 vs. 45 days, p<0.001). Secondary transplantation further demonstrated reduced functional LSC frequency, as recipients of bone marrow (BM) from GNA-treated donors survived significantly longer than controls (MllPTD/WT/Flt3ITD/ITD: MS,35 vs. 27 days, p<0.001; FLT3-WT PDX: MS,40 vs. 30 days, p<0.0001).
Although venetoclax/azacitidine (VEN/AZA) is the current standard-of-care, relapse remains a challenge due to persistent therapy-resistant LSCs. Because GNA targets NRF2-dependent mitochondrial metabolism and ferroptosis in LSCs, we hypothesized that GNA would complement VEN/AZA by eliminating residual LSCs and improving therapeutic efficacy. We demonstrate combination of GNA with VEN/AZA synergistically suppressed FAO/OXPHOS, impaired mitochondrial fusion, and enhanced ferroptosis in AML cell lines and primary AML cells, resulting in potent inhibition of cell growth (max synergy score, 26.9–39.4 in AML cell lines and 18.1–20.7 in primary AML cells). In FLT3-WT PDX models, GNA/VEN/AZA significantly reduced leukemia burden and prolonged survival compared with vehicle or either single agent in both primary (MS,60 vs. 49 [GNA], 50 [VEN/AZA], and 43 [vehicle] days; all p<0.001) and secondary transplantation models (MS,51.5 vs. 40 [GNA], 38 [VEN/AZA], and 30 [vehicle] days; all p<0.001).
In summary, we identify GNA as a novel NRF2-targeting therapeutic that disrupts mitochondrial metabolism and dynamics, promotes ferroptosis, and eliminates AML LSCs. GNA further enhances the efficacy of VEN/AZA, providing a strong preclinical rationale for clinical development of GNA-based combination therapy in AML.
DISCLOSURES
Yuying Zhu, PhD: No relevant conflicts of interest to declare.